10 open-source repos for realistic digital humans: LiveTalking, MuseTalk, Hallo and more
Shruti_0810 · x · 2026-10-14
A curated list of 10 open-source GitHub projects for making digital humans look and sound real, covering talking avatars, lip-sync, facial animation and AI portrait video:
- LiveTalking: real-time interactive avatar engine with audio-video sync, voice cloning, full-body splicing and WebRTC/RTMP streaming.
- MuseTalk: audio-driven lip-sync generation.
- Hallo: turns a portrait into a talking video via audio-driven facial animation.
- SadTalker: animates still photos with speech and head movement; now Apache 2.0 with commercial restriction removed.
- LivePortrait: expressive portrait animation from motion references.
- Wav2Lip: classic lip synchronization.
- EchoMimic: expressive talking-face generation.
A useful index for developers building AI avatar or VTuber workflows.
More from Multimodal
- Gemini 4 Argon autonomously sculpts detailed pagoda complex in Blender via multi-agent loops — DeryaTR_ · 2026-10-14
- 280 Gemini 4 Argon Creations Compiled and Open-Sourced with Prompts — yihui_indie · 2026-10-14
- ComfyUI v0.40 Ships with Qwen-Image 2.1 Turbo Support and LoRA Stack Loaders — Gremlation · 2026-10-14
- Ghost in the Shell-style AI artwork goes viral: red-dress pianist meets Motoko hologram — tetsuoai · 2026-10-14
- Magnific builds a full brand with Magnific One and Claude Opus via MCP — aziz4ai · 2026-10-14
- HeyGen's first general-purpose video model goes live on Civitai with sound in every clip — toolstelegraph · 2026-10-14